In recent times, multiple enterprises, including telecommunications operators and the Internet, have accelerated the iteration and upgrading of general large models, and launched industry large models targeting industries such as industry, finance, and transportation. Accelerate the implementation of industry oriented large models, further unleashing the industrial momentum of artificial intelligence. On November 10th, China Telecom launched an upgraded one hundred billion parameter large model, Star Semantics, to the public. According to the relevant person in charge of China Telecom, the semantic model with billions of parameters of Xingchen focuses on solving the problems of illusion and multiple rounds of logical reasoning faced by the semantic model with billions of parameters in the process of commercial implementation. The main focus is on improving the ability to generate and understand images and texts, with a 30% increase in the ability to generate Chinese image analysis. This is not an exception. Recently, domestic related companies such as Baidu, Alibaba, and iFlytek have launched upgraded versions of AI models. For example, Alibaba Cloud has released a hundred billion level parameter model called Tongyi Qianwen 2.0, which significantly improves its abilities in complex instruction comprehension, literary creation, general mathematics, knowledge memory, and other aspects. IFLYTEK has upgraded iFLYTEK's Spark Cognitive Model V3.0 and launched training on a larger parameter scale for the Spark Big Model. The development of artificial intelligence represented by large models has shown characteristics such as rapid technological innovation, strong application penetration, and fierce international competition, demonstrating a strong empowering effect. The relevant person in charge of the Ministry of Industry and Information Technology recently introduced that the scale of China's core artificial intelligence industry is constantly growing, with over 4400 enterprises, and innovative achievements such as intelligent chips, development frameworks, and universal large models are constantly emerging. At the same time, the industry model has also attracted many enterprises to increase their layout. Recently, China Mobile, in collaboration with central enterprises such as China Information Science and Technology Corporation, China Aviation Information Technology Corporation, and China Aviation Oil Corporation, released the "Nine Days · Crowd Engine Base Big Model". The base model integrates professional knowledge from eight industries, including communication, energy, steel, construction, and transportation, and will "tailor" industry models and create intelligent applications for enterprises. China Telecom has launched more than ten industry models, covering fields such as education, government services, transportation, finance, and tourism guides. Among them, the education model can reduce teacher review workload by 70%, and the grassroots governance model can increase official document writing by six times. Compared to the basic big models generated from general corpus training, industry big models have more vertically segmented datasets, professional knowledge, and process methods, which are important support for accelerating industry landing. "Zou Debao, an analyst at the Artificial Intelligence Industry Research Center of CCID Consulting, told reporters. From accelerating the development of new drugs, accurately predicting weather, to shortening factory product delivery cycles and improving office efficiency, more and more industry models are accelerating their implementation. For example, at the recently held 2023 World Internet Conference Wuzhen Summit, multiple companies showcased their new achievements in the application of big models. At the Huawei exhibition area, the Pangu Meteorological Model has attracted people's attention. It is the world's first AI model with accuracy exceeding traditional digital forecasting methods. It can complete global meteorological forecasts for the next 24 hours in just 1.4 seconds, which is more than 10000 times faster than traditional forecasting. In addition, the intelligent cabin model and intelligent cabin development tool chain developed based on the Wenxin Big Model can enable cars to have higher level intelligent driving capabilities and more powerful